v1.21.0

JuMP v1.21.0

Diff since v1.20.0

Merged pull requests:

• [docs] add tutorial on querying the basis status (#3675) (@odow)
• [docs] add section on Parameters to manual/nonlinear.md (#3679) (@odow)
• [Containers] use OrderedDict as the data structure for SparseAxisArray (#3681) (@odow)
• [docs] clarify section on automatic differentiation in nonlinear.md (#3683) (@odow)
• Add _CONSTRAINT_LIMIT_FOR_PRINTING to control printing of many constraints (#3686) (@odow)
• Update to julia-actions/cache (#3687) (@odow)
• Fix complex variables of custom coefficient type (#3691) (@blegat)
• Fix LinearAlgebra.hermitian(::AbstractJuMPScalar, ::Symbol) (#3693) (@odow)
• Fix *(::AbstractJuMPScalar{<:Real}, ::Hermitian) to return Hermitian (#3695) (@odow)
• Add the MINLP solver MAiNGO to the documentation (#3696) (@MAiNGO-github)
• [docs] add Transitioning from MATLAB tutorial (#3698) (@odow)
• [docs] add indicator trick to disjunction section (#3702) (@odow)
• [docs] clarify that JuMP can reformulate indicator constraints (#3703) (@odow)
• Add support for Hermitian PSD for conic inequality (#3705) (@blegat)
• Fix list-style for bibliographies in dark mode (#3706) (@goerz)
• Bump compat of DimensionalData (#3707) (@odow)
• [docs] improve docstrings in objective.jl (#3708) (@odow)
• [docs] standardize on '## Example' headers instead of plural (#3709) (@odow)
• [docs] improve docstrings in src/variables.jl (#3710) (@odow)
• [docs] add deprecation notes to the legacy nonlinear interface docstrings (#3711) (@odow)
• [docs] add Automatic differentiation of user-defined operators tutorial (#3713) (@odow)
• Add batched modification methods (#3716) (@joaquimg)
• [docs] update docs/packages.toml (#3717) (@odow)
• Prep for v1.21.0 (#3718) (@odow)

Closed issues:

• Documentation is too big (#3479)
• Model printing is rather verbose (#3651)
• Iterating SparseAxisArray does not preserve order (#3678)
• How to improve the speed to build a complex model? (#3688)
• Coefficients of complex variables created with a GenericModel are always Float64 (#3689)
• LinearAlgebra.hermitian is incorrect (#3692)
• *(::Real, ::Hermitian) is not hermitian (#3694)
• Cannot convert an object of type Float64 to an object of type JuMP.NonlinearExpr (#3699)
• Documentation Request: List whether a solver supports Indicator Constraints (#3701)
• Bullet point alignment in bibliography (#3704)
• Multiplication of matrix expression and variables leads to stack overflow and matmul error (#3714)
• Multiple Ranges for variables (#3715)

v1.20.0

JuMP v1.20.0

Diff since v1.19.0

Merged pull requests:

• [docs] Add Optim to list of solvers (#3624) (@blegat)
• Bump compat of DimensionalData (#3666) (@odow)
• Allow MOI.ModelLike as the optimizer instead of MOI.AbstractOptimizer (#3667) (@odow)
• Add is_solved_and_feasible (#3668) (@odow)
• [docs] add links to status and result codes (#3669) (@odow)
• Fix converting Number to GenericNonlinearExpr (#3672) (@odow)
• Prep for v1.20.0 (#3674) (@odow)
• [README] use stable link instead of version-specific (#3676) (@odow)
• [README] refactor links into table (#3677) (@odow)

Closed issues:

• NumFOCUS: GSoC 2024 Update (#3665)
• Error vcat(::NonlinearExpr, ::VariableRef, ::Float64) (#3671)
• Increase performance of SOC constraints (#3673)

v1.19.0

JuMP v1.19.0

Diff since v1.18.1

Merged pull requests:

• [docs] Add SDPLR to list of solvers (#3644) (@blegat)
• [docs] add new roadmap items (#3645) (@odow)
• [docs] fix vale version (#3650) (@odow)
• Throw informative error when parsing reference sets in macros (#3653) (@odow)
• Fix short circuting of && and || in macros (#3655) (@odow)
• Add support for modifying quadratic coefficients (#3658) (@odow)
• Tidy docstrings in src/constraints.jl (#3659) (@odow)
• Remove unneeded _affine_coefficient (#3660) (@odow)
• Move set_normalized_coefficient to src/variables.jl to better type arguments (#3661) (@odow)
• [docs] update location of Hypatia.jl (#3662) (@odow)
• Prep for v1.19.0 (#3663) (@odow)

Closed issues:

• Diagnostics request: using variable in variable container bounds (#3652)
• && and || do not short-circuit in macros (#3654)
• Add support for MOI.ScalarQuadraticCoefficientChange (#3657)

v1.18.1

JuMP v1.18.1

Diff since v1.18.0

Merged pull requests:

• Fix escaping of set kwarg in variable macro (#3647) (@odow)
• Prep for v1.18.1 (#3648) (@odow)

Closed issues:

• @variable macro fails to resolve variables in 1.18.0 (#3646)
• jump.dev is down (#3649)

v1.18.0

JuMP v1.18.0

Diff since v1.17.0

Merged pull requests:

• [docs] add DocumenterCitations (#3596) (@blegat)
• Update extension-tests workflow to include DisjunctiveProgramming (#3597) (@pulsipher)
• [docs] add DisjunctiveProgramming (#3598) (@odow)
• Rename _error to error_fn (#3600) (@odow)
• Fix error for unsupported objective sense (#3601) (@odow)
• Refactor plural macro generation code (#3603) (@odow)
• [docs] update Percival README (#3605) (@odow)
• Refactor isexpr and quot to use Meta. prefix (#3606) (@odow)
• Refactor src/macros.jl into separate files (#3607) (@odow)
• Migrate from SnoopPrecompile to PrecompileTools (#3608) (@blegat)
• Fix text/latex printing of GenericNonlinearExpr (#3609) (@odow)
• Tidy macros/@objective.jl (#3610) (@odow)
• Tidy macros/@expression.jl (#3611) (@odow)
• Tidy macros/@constraint.jl (#3612) (@odow)
• Unify handling of _finalize_macro (#3613) (@odow)
• Tidy macros/@variable.jl (#3614) (@odow)
• Remove _constraint_macro function in macros/@constraint.jl (#3615) (@odow)
• Add parse_macro_arguments to unify how we handle macro inputs (#3616) (@odow)
• [docs] clean up and clarify macro docstrings (#3617) (@odow)
• [Containers] remove unneeded prefix (#3618) (@odow)
• More refactoring of src/macros.jl (#3619) (@odow)
• [Containers] add utilities to support extensions writing macros (#3620) (@odow)
• Rename test_throws_strip and test_macro_throws for clarity (#3621) (@odow)
• [docs] update should_i_use.md to remove optimal control (#3623) (@odow)
• [docs] mention extensions in relevant tutorials (#3625) (@odow)
• Fix stdlib compat bounds in Project.toml (#3626) (@odow)
• Add missing .jl for Percival (#3628) (@blegat)
• [docs] tick off macro refactoring from roadmap (#3629) (@odow)
• [docs] add more references for DocumenterCitations.jl (#3630) (@odow)
• Fix escaping of base_name arguments in macros (#3631) (@odow)
• Fix broadcasting f-in-set in constraint macro (#3632) (@odow)
• Improve test coverage in src/macros/@variable.jl (#3633) (@odow)
• [docs] tick off constraint programming from roadmap (#3635) (@odow)
• Prep for v1.18.0 (#3636) (@odow)
• Fix modifying user-expressions in macros (#3639) (@odow)
• [docs] update to GA-4 (#3640) (@odow)
• [docs] add modulo to tips_and_tricks.jl (#3641) (@odow)
• Fix converting (Aff,Quad)Expr to NonlinearExpr (#3642) (@odow)
• Fix docs for printing changes in v1.10 (#3643) (@odow)

Closed issues:

• Improve support for constraint programming (#2227)
• Refactor JuMP macros (#3513)
• Improve LaTeX Printing for GenericNonlinearExprs (#3604)
• Performance downgrade in solving optimal control problems(OCP) (#3622)
• shorthand for GenericAffExpr{T, GenericVariableRef{T}}) (#3634)
• Why I can't read nl file with other solver (#3637)
• Constraint macro modifies expression (#3638)

v1.17.0

JuMP v1.17.0

Diff since v1.16.0

Merged pull requests:

• Add start_value, lower_bound, and upper_bound support for GenericAffExpr (#3551) (@odow)
• Add SkipModelConvertScalarSetWrapper (#3552) (@odow)
• Fix variable_ref_type for unsupported types and GenericNonlinearExpr (#3556) (@odow)
• Add errors for Matrix ± AbstractJuMPScalar (#3557) (@jd-foster)
• Improve error message for non-broadcasted addition and subtraction (#3558) (@odow)
• Add EAGO to solvers (#3560) (@DimitriAlston)
• [docs] add EAGO to packages.toml (#3561) (@odow)
• [docs] move manual section on Semicontinuous and Semiinteger variables (#3562) (@odow)
• [docs] add tutorial on piecewise linear (#3563) (@odow)
• [docs] add start_value guidance for HermitianPSDCone variables (#3564) (@odow)
• [docs] update docstrings for SOS1 and SOS2 (#3565) (@odow)
• Add Percival to list of solvers (#3567) (@blegat)
• Add Manopt to list of solvers (#3568) (@blegat)
• [docs] Classifiers tutorial (#3569) (@jd-foster)
• [docs] add section on function tracing (#3570) (@odow)
• Convert constraint starting values (#3571) (@blegat)
• Add lp_matrix_data (#3573) (@odow)
• Truncate printing of large expressions (#3575) (@odow)
• [docs] document how to support operators with several vector arguments (#3577) (@odow)
• [docs] update package compat bounds (#3578) (@odow)
• [docs] clarify indicator constraints must use variables for LHS (#3582) (@odow)
• Throw better error for non-constant variable bounds and starting value (#3583) (@odow)
• Fix error message for invalid indicator constraints (#3584) (@odow)
• Add a few missing methods for AbstractJuMPScalar to support e.g. Distances.jl (#3585) (@LebedevRI)
• [Containers] fix multi-arg eachindex (#3587) (@odow)
• Prep for v1.17.0 (#3588) (@odow)
• Fix type instability in set_objective_coefficient (#3590) (@joaquimg)
• Add ZeroOne and Integer support to lp_matrix_data (#3591) (@odow)
• [docs] add warning to SkipModelConvertScalarSetWrapper (#3592) (@odow)
• Add journal to README.md (#3593) (@joaquimg)
• [docs] Correct a small grammatical Error (#3595) (@aadya940)

Closed issues:

• Make _DoNotConvertSet Public (#3549)
• set_start_value for complex variables (#3550)
• Multiplying JuMP expression with UniformScaling absorbs the matrix (#3553)
• Provide better error message for Vector+VariableRef (#3554)
• variable_ref_type induces StackOverFlowError (#3555)
• Starting values for constant expression (#3559)
• Type instability in set_normalized_rhs (#3566)
• Add standard_form_matrix (#3572)
• Truncate printing expressions if they are very large (#3574)
• Operator with vector arguments (#3576)
• MethodError: Cannot convert an object of type NonlinearExpr to an object of type Float64 (#3579)
• MethodError: no method matching _build_indicator_constraint(<...>) (#3580)
• MethodError: no method matching isless(::Int64, ::QuadExpr) / isless(::VariableRef, ::QuadExpr) (#3581)
• [Containers] implement multi-arg eachindex (#3586)
• Failed to generate a custom sysimage for JuMP (#3594)

v1.16.0

JuMP v1.16.0

Diff since v1.15.1

Merged pull requests:

• [docs] add Octeract to supported solvers (#3527) (@odow)
• [docs] update to [email protected] (#3528) (@odow)
• [docs] add MiniZinc.jl to installation table (#3529) (@odow)
• Add := operator for Boolean satisfiability problems (#3530) (@odow)
• [docs] fix benders_decomposition.jl (#3532) (@odow)
• [docs] update getting_started_with_julia.jl (#3534) (@odow)
• Fix text/latex printing of Interval sets (#3537) (@odow)
• [docs] add Octeract to accept.txt (#3538) (@odow)
• Fix overwritten tests with duplicate function names (#3539) (@odow)
• [docs] add reference to Polyhedra.jl (#3542) (@odow)
• Add stdlib compat to Project.toml (#3544) (@odow)
• Add PolyJuMP KKT and QCQP solvers (#3545) (@blegat)
• Add MINOTAUR to solver list (#3546) (@jd-foster)
• Prep for v1.16.0 (#3547) (@odow)

Closed issues:

• 2 problems found in doc Benders Decomposition (#3531)
• Use of Σ is not consistent (#3533)
• Printing constraints with intervals escapes the bounds once too much (#3535)
• Nonlinear constraint MOI set value not converted to Float (#3540)
• Highlight Polyhedra.jl (#3541)
• MethodError: no method matching (:: LinearInterpolation {Vector {Real}, Vector {Real}, true, Real}) (:: AffExpr) occurs when DataInterpolation is called in the optimization objective (#3543)

v1.15.1

JuMP v1.15.1

Diff since v1.15.0

Merged pull requests:

• [docs] update to [email protected] (#3501) (@odow)
• Throw error for add_to_expression with GenericNonlinearExpr (#3506) (@odow)
• [docs] add links to the various data files in tutorials (#3512) (@odow)
• Fix constraint tags for broadcast constraints (#3515) (@odow)
• Fix typo in the objective of the TSP lazy constraints example. (#3516) (@soppera)
• Fix invalid call to InexactError in MA.scaling (#3518) (@odow)
• Fix various JET errors (#3519) (@odow)
• Add explanatory error message for VariableNotOwned errors (#3520) (@odow)
• Fix unary min and max operators (#3522) (@odow)
• Fix constructing an array of Parameter variables (#3524) (@odow)
• Prep for v1.15.1 (#3525) (@odow)
• Fix documentation.yml GitHub action (#3526) (@odow)

Closed issues:

• Constraints with few terms are (relatively) slow (#2817)
• Extend add_to_expression! for NonlinearExpr (#3505)
• Drop zeros for nonlinear functions (#3510)
• Update location of 'database called multi.sqlite' for tutorial 'The multi-commodity flow problem' (#3511)
• Constraints with broadcasting operators hang with tags (#3514)
• Make VariableNotOwned error more explanatory (#3517)
• More Parameter woes (#3523)

v1.15.0

JuMP v1.15.0

Diff since v1.14.1

Merged pull requests:

• Fix integer powers to return appropriate type instead of QuadExpr (#3474) (@odow)
• Fix errors for LinearAlgebra operators of JuMP arrays (#3476) (@odow)
• Prep for v1.14.1 (#3481) (@odow)
• Fix promotion rules for GenericNonlinearExpr (#3483) (@odow)
• [docs] update to latest version of Documenter.jl (#3484) (@odow)
• Prep for v1.15.0 (#3485) (@odow)
• Fix MA.Zero in MA.add_mul implementation of GenericNonlinearExpr (#3487) (@odow)
• Fix MA.add_mul for GenericNonlinearExpr arguments (#3488) (@odow)
• Refactor GenericNonlinearExpr and NonlinearOperator constructors (#3489) (@odow)
• [docs] update performance_tips tutorial (#3490) (@odow)
• [docs] add multi-commodity network example (#3491) (@odow)
• [docs] re-order the linear program tutorials (#3492) (@odow)
• Fix error messages and documention for variable bound functions (#3494) (@odow)
• [docs] update package versions (#3495) (@odow)
• [docs] move literate footer to header (#3497) (@odow)
• Use let model=model in variable macro to improve type stability (#3500) (@odow)
• [docs] fix warning block (#3503) (@odow)
• [docs] cross nonlinear update off roadmap (#3504) (@odow)
• Update GitHub action versions (#3507) (@odow)
• Add support for min and max operators in GenericNonlinearExpr (#3509) (@odow)

Closed issues:

• Fast resolves for NLP (#1185)
• Consider changing x^1 to return x (#3454)
• StackOverflow for norm of vector of variables (#3467)
• Documentation: links to API elements broken in Changelog (#3477)
• The Vector{AbstractJuMPScalar} problem (#3482)
• MA.Zero in nonlinear expressions (#3486)
• missleading error message: fix_value for unfixed variable (#3493)
• Links to the source code for tutorials download instead of displaying on GitHub (#3496)
• @variable declarations returning empty container should have VariableRef element type (#3499)
• New nonlinear expression stuff doesn't support min properly (#3508)

v1.14.1

JuMP v1.14.1

Diff since v1.14.0

Closed issues:

• JuMP 1.14.0 fails to precompile on Julia 1.6.0/1 (#3461)
• StackOverflow with operate!! and NonlinearExpr (#3469)
• Nonlinear interfaces are not well distinguished in the docs (#3471)

Merged pull requests:

• [Nonlinear] add new nonlinear interface (#3106) (@odow)
• [docs] add GAMS to the documentation (#3357) (@odow)
• Add upper_triangle and docs for {Log,Root}DetCone{Triangle,Square} (#3456) (@odow)
• [docs] update The knapsack problem example (#3459) (@odow)
• [docs] update The diet problem (#3460) (@odow)
• [docs] update Multi-objective knapsack (#3462) (@odow)
• [docs] update Rocket Control tutorial (#3463) (@odow)
• Fix use of @nospecialize with default arguments (#3464) (@odow)
• [docs] update Column generation tutorial (#3465) (@odow)
• [docs] add two-stage stochastic program tutorial (#3466) (@odow)
• Change argument order of NonlinearOperator (#3468) (@blegat)
• [docs] distinguish Nonlinear Modeling sections in documentation (#3472) (@odow)
• Fix vector of parameters and add documentation (#3473) (@odow)
• Fix performance of MA.add_mul for NonlinearExpr (#3475) (@odow)
• Revert to an older version of Documenter.jl (#3478) (@odow)
• Prep for v1.14.1 (#3480) (@odow)